import pandas as pd
import numpy as np
import seaborn as sns
import pylab as py

%pylab inline

Populating the interactive namespace from numpy and matplotlib

train=pd.read_csv('train.csv')

train

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
28826	customer_id_6265	60	retired	married	professional.course	unknown	no	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0	0
28827	customer_id_11284	39	management	married	university.degree	no	no	no	telephone	jun	...	1	999	0	nonexistent	1.43	47.2325	-29.26	5.963	5228.1	0
28828	customer_id_38158	37	admin.	married	high.school	no	yes	no	cellular	oct	...	1	4	1	success	-3.37	46.2155	-18.83	1.756	5017.5	1
28829	customer_id_860	42	management	married	university.degree	no	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.858	5191.0	0
28830	customer_id_15795	31	admin.	single	university.degree	no	yes	no	cellular	jul	...	2	999	0	nonexistent	1.43	46.9590	-29.89	5.962	5228.1	0

28831 rows × 22 columns

test=pd.read_csv('test.csv')

test.sample(10)

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	duration	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed
5909	customer_id_20196	34	technician	married	university.degree	no	no	yes	cellular	aug	...	5.516667	5	999	0	nonexistent	1.43	46.7220	-25.27	5.967	5228.1
5960	customer_id_32783	35	unemployed	married	basic.9y	no	yes	yes	cellular	may	...	0.350000	6	999	0	nonexistent	-1.77	46.4465	-32.34	2.301	5099.1
766	customer_id_5323	49	services	married	professional.course	no	yes	no	telephone	may	...	3.766667	2	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0
8331	customer_id_2075	58	housemaid	divorced	basic.4y	unknown	no	no	telephone	may	...	1.433333	1	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0
5875	customer_id_19647	54	technician	divorced	high.school	no	yes	no	cellular	aug	...	1.683333	1	999	0	nonexistent	1.43	46.7220	-25.27	5.970	5228.1
6834	customer_id_18843	60	blue-collar	married	basic.9y	no	no	no	cellular	aug	...	8.300000	1	999	0	nonexistent	1.43	46.7220	-25.27	5.972	5228.1
2109	customer_id_11836	43	entrepreneur	married	university.degree	unknown	yes	no	telephone	jun	...	3.600000	2	999	0	nonexistent	1.43	47.2325	-29.26	5.961	5228.1
2450	customer_id_2877	37	technician	married	professional.course	no	no	no	telephone	may	...	2.683333	2	999	0	nonexistent	1.13	46.9970	-25.48	5.861	5191.0
7542	customer_id_8889	40	services	married	high.school	unknown	no	no	telephone	jun	...	6.983333	1	999	0	nonexistent	1.43	47.2325	-29.26	5.868	5228.1
11505	customer_id_4842	31	blue-collar	married	basic.9y	no	yes	no	telephone	may	...	8.200000	1	999	0	nonexistent	1.13	46.9970	-25.48	5.860	5191.0

10 rows × 21 columns

Target

target=[]
for i in train.columns:
    if i not in test.columns:
        target.append(i)
        print(target)

['subscribed']

train.head(10)

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
5	customer_id_24606	48	entrepreneur	married	university.degree	no	yes	yes	cellular	nov	...	2	999	1	failure	-0.07	46.6000	-29.40	5.193	5195.8	0
6	customer_id_13094	34	blue-collar	single	basic.4y	unknown	no	no	cellular	jul	...	2	999	0	nonexistent	1.43	46.9590	-29.89	5.964	5228.1	0
7	customer_id_36912	34	technician	single	professional.course	no	no	no	cellular	jun	...	1	999	0	nonexistent	-2.87	46.4815	-28.56	2.217	5076.2	0
8	customer_id_27834	39	unemployed	single	university.degree	no	yes	no	cellular	mar	...	1	999	0	nonexistent	-1.77	46.4215	-35.00	2.642	5099.1	0
9	customer_id_9302	40	technician	single	professional.course	no	unknown	unknown	telephone	jun	...	3	999	0	nonexistent	1.43	47.2325	-29.26	5.969	5228.1	0

10 rows × 22 columns

import cufflinks as cf
from pandas_profiling import ProfileReport

train.isnull()

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
1	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
2	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
3	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
4	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
28826	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
28827	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
28828	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
28829	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False
28830	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False

28831 rows × 22 columns

train.isnull().sum()

customer_id       0
age               0
job               0
marital           0
education         0
default           0
housing           0
loan              0
contact           0
month             0
day_of_week       0
duration          0
campaign          0
pdays             0
previous          0
poutcome          0
emp_var_rate      0
cons_price_idx    0
cons_conf_idx     0
euribor3m         0
nr_employed       0
subscribed        0
dtype: int64

train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 28831 entries, 0 to 28830
Data columns (total 22 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   customer_id     28831 non-null  object 
 1   age             28831 non-null  int64  
 2   job             28831 non-null  object 
 3   marital         28831 non-null  object 
 4   education       28831 non-null  object 
 5   default         28831 non-null  object 
 6   housing         28831 non-null  object 
 7   loan            28831 non-null  object 
 8   contact         28831 non-null  object 
 9   month           28831 non-null  object 
 10  day_of_week     28831 non-null  object 
 11  duration        28831 non-null  float64
 12  campaign        28831 non-null  int64  
 13  pdays           28831 non-null  int64  
 14  previous        28831 non-null  int64  
 15  poutcome        28831 non-null  object 
 16  emp_var_rate    28831 non-null  float64
 17  cons_price_idx  28831 non-null  float64
 18  cons_conf_idx   28831 non-null  float64
 19  euribor3m       28831 non-null  float64
 20  nr_employed     28831 non-null  float64
 21  subscribed      28831 non-null  int64  
dtypes: float64(6), int64(5), object(11)
memory usage: 4.8+ MB

train.describe()

	age	duration	campaign	pdays	previous	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
count	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000	28831.000000
mean	42.011203	4.297919	2.575769	963.215844	0.172592	0.113202	46.788632	-28.360564	4.623599	5167.011880	0.112761
std	10.450128	4.336882	2.752303	185.077567	0.494338	1.570978	0.289847	3.244405	1.735202	72.542598	0.316305
min	19.000000	0.000000	1.000000	0.000000	0.000000	-3.370000	46.100500	-35.560000	1.636000	4963.600000	0.000000
25%	34.000000	1.700000	1.000000	999.000000	0.000000	-1.770000	46.537500	-29.890000	2.346000	5099.100000	0.000000
50%	40.000000	3.000000	2.000000	999.000000	0.000000	1.130000	46.874500	-29.260000	5.859000	5191.000000	0.000000
75%	49.000000	5.300000	3.000000	999.000000	0.000000	1.430000	46.997000	-25.480000	5.963000	5228.100000	0.000000
max	100.000000	81.966667	43.000000	999.000000	7.000000	1.430000	47.383500	-18.830000	6.047000	5228.100000	1.000000

fig=py.figure(figsize=(10,8))
sns.set(style='darkgrid')
sns.heatmap(train.isnull(),yticklabels=False)

<AxesSubplot:>

train.profile_report()

train.columns

Index(['customer_id', 'age', 'job', 'marital', 'education', 'default',
       'housing', 'loan', 'contact', 'month', 'day_of_week', 'duration',
       'campaign', 'pdays', 'previous', 'poutcome', 'emp_var_rate',
       'cons_price_idx', 'cons_conf_idx', 'euribor3m', 'nr_employed',
       'subscribed'],
      dtype='object')

train.head()

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0

5 rows × 22 columns

fig=py.figure(figsize=(10,8))
sns.boxplot(y='age',x='job',data=train)
py.tight_layout()
py.xticks(rotation=90)

(array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11]),
 [Text(0, 0, 'admin.'),
  Text(1, 0, 'technician'),
  Text(2, 0, 'blue-collar'),
  Text(3, 0, 'services'),
  Text(4, 0, 'entrepreneur'),
  Text(5, 0, 'unemployed'),
  Text(6, 0, 'housemaid'),
  Text(7, 0, 'management'),
  Text(8, 0, 'unknown'),
  Text(9, 0, 'self-employed'),
  Text(10, 0, 'retired'),
  Text(11, 0, 'student')])

train['job'].unique()

array(['admin.', 'technician', 'blue-collar', 'services', 'entrepreneur',
       'unemployed', 'housemaid', 'management', 'unknown',
       'self-employed', 'retired', 'student'], dtype=object)

fig=py.figure(figsize=(10,8))
train[train['housing']=='yes']['cons_price_idx'].plot(kind='hist',bins=30,label='yes')
train[train['housing']=='no']['cons_price_idx'].plot(kind='hist',bins=30,label='no')
py.legend()

<matplotlib.legend.Legend at 0x1f381de35e0>

train.groupby(by='marital').count()['education'].plot()

<AxesSubplot:xlabel='marital'>

train.head()

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0

5 rows × 22 columns

train

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
28826	customer_id_6265	60	retired	married	professional.course	unknown	no	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0	0
28827	customer_id_11284	39	management	married	university.degree	no	no	no	telephone	jun	...	1	999	0	nonexistent	1.43	47.2325	-29.26	5.963	5228.1	0
28828	customer_id_38158	37	admin.	married	high.school	no	yes	no	cellular	oct	...	1	4	1	success	-3.37	46.2155	-18.83	1.756	5017.5	1
28829	customer_id_860	42	management	married	university.degree	no	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.858	5191.0	0
28830	customer_id_15795	31	admin.	single	university.degree	no	yes	no	cellular	jul	...	2	999	0	nonexistent	1.43	46.9590	-29.89	5.962	5228.1	0

28831 rows × 22 columns

test.head()

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	duration	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed
0	customer_id_32884	59	technician	married	high.school	no	no	yes	cellular	may	...	6.183333	1	999	1	failure	-1.77	46.4465	-32.34	2.301	5099.1
1	customer_id_3169	57	unknown	married	unknown	unknown	yes	no	telephone	may	...	4.750000	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0
2	customer_id_32206	35	blue-collar	married	basic.9y	no	no	no	cellular	may	...	0.866667	1	999	1	failure	-1.77	46.4465	-32.34	2.315	5099.1
3	customer_id_9403	38	admin.	married	high.school	no	no	no	telephone	jun	...	5.916667	4	999	0	nonexistent	1.43	47.2325	-29.26	5.969	5228.1
4	customer_id_14020	29	housemaid	married	high.school	no	yes	no	cellular	jul	...	3.150000	2	999	0	nonexistent	1.43	46.9590	-29.89	5.965	5228.1

5 rows × 21 columns

train_id=train.select_dtypes(include=['number'])

train_id

	age	duration	campaign	pdays	previous	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	31	1.283333	3	999	1	-2.97	46.3565	-23.10	1.711	5023.5	0
1	31	0.200000	4	999	0	-1.77	46.4465	-32.34	2.252	5099.1	0
2	47	4.616667	2	999	0	1.13	46.9970	-25.48	5.862	5191.0	0
3	36	1.166667	1	999	1	-1.77	46.4465	-32.34	2.329	5099.1	0
4	34	19.683333	9	999	0	-1.77	46.4465	-32.34	2.252	5099.1	0
...	...	...	...	...	...	...	...	...	...	...	...
28826	60	7.116667	2	999	0	1.13	46.9970	-25.48	5.859	5191.0	0
28827	39	4.800000	1	999	0	1.43	47.2325	-29.26	5.963	5228.1	0
28828	37	3.233333	1	4	1	-3.37	46.2155	-18.83	1.756	5017.5	1
28829	42	4.916667	2	999	0	1.13	46.9970	-25.48	5.858	5191.0	0
28830	31	5.550000	2	999	0	1.43	46.9590	-29.89	5.962	5228.1	0

28831 rows × 11 columns

from sklearn.linear_model import LogisticRegression

X=train_id.drop(target,axis=1)
y=train_id['subscribed']

To verify the most important features using Extratreeclassifier model

from sklearn.ensemble import ExtraTreesClassifier

model=ExtraTreesClassifier()

model.fit(X,y)

ExtraTreesClassifier()

model.feature_importances_

array([0.15837812, 0.39500592, 0.07265032, 0.07162169, 0.02224527,
       0.02634433, 0.02122306, 0.02618026, 0.15328945, 0.05306158])

cdf=pd.DataFrame(model.feature_importances_,index=X.columns,columns=['features'])

fig=py.figure(figsize=(15,10))
cdf.plot(kind='barh',figsize=(10,10))

<AxesSubplot:>

<Figure size 1080x720 with 0 Axes>

model_log=LogisticRegression(max_iter=500)

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.42859, random_state=101)

model_log.fit(X_train,y_train)

LogisticRegression(max_iter=500)

predict=model_log.predict(X_test)

predict

array([0, 0, 0, ..., 0, 1, 0], dtype=int64)

from sklearn.metrics import classification_report , confusion_matrix,accuracy_score

print(classification_report(y_test,predict))

              precision    recall  f1-score   support

           0       0.92      0.97      0.95     10929
           1       0.65      0.35      0.46      1428

    accuracy                           0.90     12357
   macro avg       0.78      0.66      0.70     12357
weighted avg       0.89      0.90      0.89     12357

print(confusion_matrix(y_test,predict))

[[10651   278]
 [  922   506]]

accuracy_score(y_test,predict)

0.902889050740471

sns.heatmap(confusion_matrix(y_test,predict),annot=True,cmap='viridis')

<AxesSubplot:>

print(len(predict))
print('\n')
len(sample)

12357

12357

sample=pd.read_csv('sample_submission.csv')

sample.head()

	customer_id	subscribed
0	customer_id_32884	1
1	customer_id_3169	1
2	customer_id_32206	1
3	customer_id_9403	1
4	customer_id_14020	1

len(sample)

12357

len(predict)

12357

sample=sample.copy()

sample['subscribed']=predict

sample.to_csv('firstai.csv',index=False)

pd.read_csv('firstai.csv')

	customer_id	subscribed
0	customer_id_32884	0
1	customer_id_3169	0
2	customer_id_32206	0
3	customer_id_9403	0
4	customer_id_14020	0
...	...	...
12352	customer_id_15908	0
12353	customer_id_28222	0
12354	customer_id_14194	0
12355	customer_id_19764	1
12356	customer_id_26052	0

12357 rows × 2 columns

from sklearn.svm import SVC

model_sv=SVC()

model_sv.fit(X_train,y_train)

SVC()

predict_sv=model_sv.predict(X_test)

print(classification_report(y_test,predict_sv))

              precision    recall  f1-score   support

           0       0.90      0.98      0.94     10929
           1       0.63      0.20      0.30      1428

    accuracy                           0.89     12357
   macro avg       0.76      0.59      0.62     12357
weighted avg       0.87      0.89      0.87     12357

print(confusion_matrix(y_test,predict))

[[10651   278]
 [  922   506]]

accuracy_score(y_test,predict)

0.902889050740471

sample['subscribed']=predict_sv

sample.to_csv('seconda1.csv',index=False)

from sklearn.ensemble import RandomForestClassifier
model_random=RandomForestClassifier(n_estimators=200)

model_random.fit(X_train,y_train)

RandomForestClassifier(n_estimators=200)

model_predict=model_random.predict(X_test)

model_predict

array([0, 0, 0, ..., 0, 1, 0], dtype=int64)

print(confusion_matrix(y_test,model_predict))
print('\n')
print(classification_report(y_test,model_predict))

[[10538   391]
 [  718   710]]


              precision    recall  f1-score   support

           0       0.94      0.96      0.95     10929
           1       0.64      0.50      0.56      1428

    accuracy                           0.91     12357
   macro avg       0.79      0.73      0.76     12357
weighted avg       0.90      0.91      0.91     12357

accuracy_score(y_test,predict)

0.902889050740471

sample['subscribed']=model_predict

sample.to_csv('thirdai.csv',index=False)

train

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
28826	customer_id_6265	60	retired	married	professional.course	unknown	no	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0	0
28827	customer_id_11284	39	management	married	university.degree	no	no	no	telephone	jun	...	1	999	0	nonexistent	1.43	47.2325	-29.26	5.963	5228.1	0
28828	customer_id_38158	37	admin.	married	high.school	no	yes	no	cellular	oct	...	1	4	1	success	-3.37	46.2155	-18.83	1.756	5017.5	1
28829	customer_id_860	42	management	married	university.degree	no	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.858	5191.0	0
28830	customer_id_15795	31	admin.	single	university.degree	no	yes	no	cellular	jul	...	2	999	0	nonexistent	1.43	46.9590	-29.89	5.962	5228.1	0

28831 rows × 22 columns

train_1=pd.get_dummies(train,drop_first=True)

train_1.head()

	age	duration	campaign	pdays	previous	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	...	month_may	month_nov	month_oct	month_sep	day_of_week_mon	day_of_week_thu	day_of_week_tue	day_of_week_wed	poutcome_nonexistent	poutcome_success
0	31	1.283333	3	999	1	-2.97	46.3565	-23.10	1.711	5023.5	...	0	0	0	0	1	0	0	0	0	0
1	31	0.200000	4	999	0	-1.77	46.4465	-32.34	2.252	5099.1	...	1	0	0	0	0	0	0	0	1	0
2	47	4.616667	2	999	0	1.13	46.9970	-25.48	5.862	5191.0	...	1	0	0	0	0	1	0	0	1	0
3	36	1.166667	1	999	1	-1.77	46.4465	-32.34	2.329	5099.1	...	1	0	0	0	0	1	0	0	0	0
4	34	19.683333	9	999	0	-1.77	46.4465	-32.34	2.252	5099.1	...	1	0	0	0	0	0	0	0	1	0

5 rows × 28884 columns

X=train_1.drop(target,axis=1)

y=train_1['subscribed']

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.42859, random_state=102)

from sklearn.linear_model import LogisticRegression

log_reg=LogisticRegression(max_iter=1000)

cdf=train.drop('customer_id',axis=1)

train_model=train.drop(['customer_id'],axis=1)

train_model=pd.get_dummies(train_model)

X=train_model.drop(target,axis=1)
y=train_model['subscribed']

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.42859, random_state=102)

train

	customer_id	age	job	marital	education	default	housing	loan	contact	month	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	customer_id_39075	31	admin.	married	university.degree	no	no	no	cellular	dec	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	customer_id_34855	31	technician	single	university.degree	no	no	no	telephone	may	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	customer_id_7107	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	customer_id_31614	36	services	married	university.degree	no	no	no	cellular	may	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	customer_id_34878	34	admin.	single	high.school	no	no	no	cellular	may	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
28826	customer_id_6265	60	retired	married	professional.course	unknown	no	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.859	5191.0	0
28827	customer_id_11284	39	management	married	university.degree	no	no	no	telephone	jun	...	1	999	0	nonexistent	1.43	47.2325	-29.26	5.963	5228.1	0
28828	customer_id_38158	37	admin.	married	high.school	no	yes	no	cellular	oct	...	1	4	1	success	-3.37	46.2155	-18.83	1.756	5017.5	1
28829	customer_id_860	42	management	married	university.degree	no	yes	no	telephone	may	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.858	5191.0	0
28830	customer_id_15795	31	admin.	single	university.degree	no	yes	no	cellular	jul	...	2	999	0	nonexistent	1.43	46.9590	-29.89	5.962	5228.1	0

28831 rows × 22 columns

cdf.head()

	age	job	marital	education	default	housing	loan	contact	month	day_of_week	...	campaign	pdays	previous	poutcome	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	subscribed
0	31	admin.	married	university.degree	no	no	no	cellular	dec	mon	...	3	999	1	failure	-2.97	46.3565	-23.10	1.711	5023.5	0
1	31	technician	single	university.degree	no	no	no	telephone	may	fri	...	4	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0
2	47	blue-collar	married	basic.6y	unknown	yes	no	telephone	may	thu	...	2	999	0	nonexistent	1.13	46.9970	-25.48	5.862	5191.0	0
3	36	services	married	university.degree	no	no	no	cellular	may	thu	...	1	999	1	failure	-1.77	46.4465	-32.34	2.329	5099.1	0
4	34	admin.	single	high.school	no	no	no	cellular	may	fri	...	9	999	0	nonexistent	-1.77	46.4465	-32.34	2.252	5099.1	0

5 rows × 21 columns

cdf=pd.get_dummies(cdf,drop_first=1)

cdf.head()

	age	duration	campaign	pdays	previous	emp_var_rate	cons_price_idx	cons_conf_idx	euribor3m	nr_employed	...	month_may	month_nov	month_oct	month_sep	day_of_week_mon	day_of_week_thu	day_of_week_tue	day_of_week_wed	poutcome_nonexistent	poutcome_success
0	31	1.283333	3	999	1	-2.97	46.3565	-23.10	1.711	5023.5	...	0	0	0	0	1	0	0	0	0	0
1	31	0.200000	4	999	0	-1.77	46.4465	-32.34	2.252	5099.1	...	1	0	0	0	0	0	0	0	1	0
2	47	4.616667	2	999	0	1.13	46.9970	-25.48	5.862	5191.0	...	1	0	0	0	0	1	0	0	1	0
3	36	1.166667	1	999	1	-1.77	46.4465	-32.34	2.329	5099.1	...	1	0	0	0	0	1	0	0	0	0
4	34	19.683333	9	999	0	-1.77	46.4465	-32.34	2.252	5099.1	...	1	0	0	0	0	0	0	0	1	0

5 rows × 54 columns

log_reg.fit(X_train,y_train)

LogisticRegression(max_iter=1000)

logpredict=log_reg.predict(X_test)

logpredict

array([0, 1, 0, ..., 0, 0, 0], dtype=int64)

len(logpredict)

12357

print(classification_report(y_test,logpredict))
print('\n')
print(confusion_matrix(y_test,logpredict))

              precision    recall  f1-score   support

           0       0.93      0.97      0.95     10970
           1       0.66      0.42      0.51      1387

    accuracy                           0.91     12357
   macro avg       0.79      0.70      0.73     12357
weighted avg       0.90      0.91      0.90     12357



[[10667   303]
 [  805   582]]

accuracy_score(y_test,logpredict)
sample=sample.copy()
sample['subscribed']=logpredict
sample

	customer_id	subscribed
0	customer_id_32884	0
1	customer_id_3169	1
2	customer_id_32206	0
3	customer_id_9403	0
4	customer_id_14020	0
...	...	...
12352	customer_id_15908	0
12353	customer_id_28222	0
12354	customer_id_14194	0
12355	customer_id_19764	0
12356	customer_id_26052	0

12357 rows × 2 columns

sample.to_csv('fourthai.csv',index=False)
from sklearn.ensemble import RandomForestClassifier
X=cdf.drop(target,axis=1)
y=cdf['subscribed']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.569899, random_state=102)

rdf=RandomForestClassifier()

rdf.fit(X_train,y_train)

RandomForestClassifier()

rdf_predict=rdf.predict(X_train)

len(rdf_predict)

12400

rd=pd.DataFrame(rdf_predict,columns=target)

rdf=rd[:len(sample)]

sample=sample.copy()
sample['subscribed']=rdf

sample.to_csv('sixthai.csv',index=False)

rdf

	subscribed
0	0
1	0
2	0
3	0
4	1
...	...
12352	0
12353	0
12354	0
12355	0
12356	1

12357 rows × 1 columns